An Electrospray Ionization Mass Spectrometry
Investigation of 1-Anilino-8-NaphthaleneSulfonate (ANS) Binding to Proteins
Soumya S. Ray, S. Kumar Singh, and P. Balaram
Molecular Biophysics Unit, Indian Institute of Science, Bangalore 560012, India
The binding of 1-anilino-8-naphthalene-sulfonic acid (ANS) to various globular proteins at
acidic pH has been investigated by electrospray ionization mass spectrometry (ESI-MS).
Maximal ANS binding is observed in the pH range 3–5. As many as seven species of
dye-bound complexes are detected for myoglobin. Similar studies were carried out with
cytochrome c, carbonic anhydrase, triosephosphate isomerase, lysozyme, ␣-lactalbumin, and
bovine pancreatic trypsin inhibitor (BPTI). Strong ANS binding was observed wherever
molten globule states were postulated in solution. ANS binding is not observed for lysozyme
and BPTI, which have tightly folded structures in the native form. ␣-Lactalbumin, which is
structurally related to lysozyme but forms a molten globule at acidic pH, exhibited ANS
binding. Reduction of disulfide bonds in these proteins leads to the detection of ANS binding
even at neutral pH. Binding was suppressed at very low pH (⬍2.5), presumably due to
neutralization of the charge on the sulfonate moiety. The distribution of the relative intensities
of the protein bound ANS species varies with the charge state, suggesting heterogeneity of gas
phase conformations. The binding strength of these complexes was qualitatively estimated by
dissociating them using enhanced nozzle skimmer potentials. The skimmer voltages also
affected the lower and higher charge states of these complexes in a different manner.
E
lectrospray ionization mass spectrometry (ESIMS) provides a powerful means to study noncovalent interactions in the gas phase [1–3]. The
mild ionization procedure permits protein molecules to
retain “conformational memory” of their solution state
structures in the gas phase [4 –9]. Different folded states
in solution produce different charge states in the
ESI-MS spectrum. ESI-MS has been used extensively to
establish binding stoichiometries for many protein–
ligand and protein–protein complexes. We have recently demonstrated in a preliminary report [10], that
ESI-MS can be used to detect noncovalent complexes of
the widely used fluorescent dye 1-anilino-8-naphthalene-sulfonate (ANS) [11, 12] to proteins in the gas
phase. An important issue that has been addressed in
several studies is the correlation between protein conformation in solution and the nature of the structural
states detected in the gas phase [6, 13–19]. Although
hydrophobic interactions are generally major contributors to the stabilization of folded structures and protein–ligand complexes in aqueous solutions, it is clear
that they will become less important in the gas phase.
On the contrary, electrostatic effects and hydrogen
bonds, which appear less dominant in water because of
Address reprint requests to Prof. P. Balaram, Molecular Biophysics Unit,
Indian Institute of Science, Bangalore 560012, India. E-mail:
pb@mbu.iisc.ernet.in
solvent competition, are likely to assume importance in
the gas phase.
ANS is widely used as a hydrophobic probe to study
solution state folding behavior. Completely folded and
unfolded states of proteins often do not bind ANS with
appreciable affinity. Binding is maximal under conditions where they exist in partially unfolded “molten
globule” states, which are known to provide exposed
hydrophobic sites for ANS binding [12]. Although ANS
binding to partially unfolded states has been widely
investigated by optical spectroscopic methods, there is
comparatively little information regarding the stoichiometry of binding. Recent studies suggest that ANS
binds nonspecifically to hydrophobic sites and the
interaction is mediated by formation of ion pairs [20,
21].
In this study, we analyze the interactions of ANS
with diverse proteins using ESI-MS to detect gas phase
complexes when solutions of protein and dye are electrosprayed under a variety of conditions.
Experimental Procedures
Chemicals
Myoglobin (horse heart), cytochrome c (bovine), lysozyme (hen egg white), ␣-lactalbumin (bovine), carbonic anhydrase (bovine), and BPTI (bovine) were
obtained from Sigma (St. Louis, MO) and were used
without further purification. Plasmodium falciparum triosephosphate isomerase was expressed in E. coli and
purified as described before [22]. The ANS ammonium
salt (Sigma) was recrystallized from ethanol before use.
Unfolding and ANS Binding
pH dependent unfolding was carried out using 25 mM
ammonium acetate buffer and the pH adjusted using
glacial acetic acid and ammonia solution. Each protein
(15 ␮M) was incubated with ANS (100 ␮M) at the
desired pH for 30 min at 25 °C prior to mass spectral
analysis.
Reduction of Disulfide Containing Proteins
Lysozyme, BPTI, and ␣-lactalbumin (50 ␮M) were reduced by adding 100 mM DTT and heating at 45 °C for
1 h. Excess DTT was removed by passing the solution
through a G10 column. Protein samples were stored in
an inert atmosphere and checked by ESI-MS to confirm
complete reduction.
Mass Spectrometry
ESI-MS was carried out on a Hewlett Packard (model
HP-1100) electrospray mass spectrometer. Ammonium
acetate buffer adjusted to the required pH was used as
the mobile phase at a flow rate of 10 ␮L/min. Approximately 150 pmol of sample was injected per run.
Electrospray was carried out using a capillary with an
internal diameter of 0.1 mm. The tip was held at 5000 V
in positive ion mode. Nebulization was assisted by N2
gas (99.8%) at a flow rate of 10 L/min. The spray
chamber was held at 200 °C. The ion optics zone was
optimized for maximal ion transmission. The instrument has two variable skimmers. Skimmer 2 was held
constant at 10 V, and skimmer 1 (declustering potential)
was adjusted for optimal detection. Maximal binding
was detected when skimmer 1 was held at 30 V (⌬V ⫽
20 V). Data was acquired across mass range 500 –3000
m/z with a quadrupole cycle time of 3 s. The spectrometer was tuned using five calibration standards provided by the manufacturer. Data processing was done
using the deconvolution module of the Chemstation
software. A minimum of five peaks well above the
baseline was used for molecular mass determination.
Results and Discussion
Gas Phase Detection of ANS–Protein Complexes
The binding of ANS to proteins, in principle, should
lead to increase in mass by 299 Da for each bound anion
of ANS. In the ESI-MS spectrum, protein-dye adducts
can be detected as satellite charge states around each
protein charge state. The various charge states of a
protein may show different affinities for ANS, leading
to variation in the intensity of the ANS satellite peaks.
The number and the distribution of molecular masses
obtained upon processing (deconvoluting) the raw
spectrum provides a qualitative measure of bound
moieties that survive the desolvation/ionization process.
Myoglobin
Myoglobin is a globular all-helical protein with a molecular mass of 16,951 Da (for the apo form) and binds
heme (616 Da) in its native state. Extensive mass spectral studies of myoglobin have focused on refolding
behavior [23], H–D exchange [24], gas phase stability of
the heme bound complex [25], proton transfer reactivities [26], and sensitivity to proteolysis [27]. Tsui et al.,
were able to detect an intermediate during myoglobin
unfolding using quench flow ESI-MS [28].
Under mild electrospray source conditions and a
solution pH of 2.0, myoglobin ions are highly charged
showing a charge state distribution from ⫹8 to ⫹30
centered at ⫹19 (Figure 1). Heme binding is not observed. At low pH, the prosthetic group is known to
dissociate due to titration of the heme binding histidine
residues. At pH 3.0, four distinct sets of charge states
appear, which deconvolute to masses of 16,951 Da,
17,250 Da (apomyoglobin ⫹ 1ANS), 17,548 Da (apomyoglobin ⫹ 2ANS), 17,849 Da (apomyoglobin ⫹
3ANS), and 18,146 Da (myoglobin ⫹ 4ANS). At pH 4.0,
there is a very low population of free apomyoglobin
and most of it is bound to one or more ANS molecules.
Some of the protein now binds to heme and yields a
molecular mass of 17,567 Da (holomyoglobin). It is
interesting to note that a new species having a mass of
17,867 Da corresponding to holomyoglobin bound to
one ANS was detected. This species has not been
detected previously because the heme bound to myoglobin interferes in optical spectroscopic studies of ANS
binding. At pH 6.0, only holomyoglobin is observed
with a very small proportion of apomyoglobin still
visible. The spectrum at pH 6.0 has fewer charge states,
suggesting a considerable compaction of the molecule
upon heme binding (Figure 1). At pH 2.0 the ESI-MS
spectra do not reveal any bound ANS. This is likely to
be due to titration of the sulfonic acid group of ANS,
resulting in a loss of electrostatic contributions to binding. This feature is observed in all the proteins studied.
These data suggest, that between pH 4.0 and 3.0,
dissociation of heme is almost complete and multiple
ANS molecules bind to apomyoglobin. The observation
is consistent with the finding that removal of heme
leads to the formation of a “molten globule like” floppy
structure of apomyoglobin as reported by others using
2-D NMR [29]. It is noteworthy, that the single ANS
bound species is much more intense than apomyoglobin peak itself under these electrospray conditions.
Between pH 6.0 and 4.0, the heme bound form of
myoglobin undergoes acid expansion resulting in exposure of a new hydrophobic binding site for ANS. The
Figure 1. ESI mass spectrum of myoglobin (left) and triosephosphate isomerase (pfTIM) (right) at
different pH values after incubation with excess ANS. The masses derived by deconvolution are
shown to the right of the spectra. For spectra showing multiple species, the mass values obtained by
deconvolution are indicated. In the case of myoglobin at pH 4.0, the species A* and B* correspond to
holomyoglobin and holomyoglobin ⫹ 1ANS, respectively.
observation of a species at pH 4.0, which corresponds to
holomyoglobin with a single bound ANS molecule,
suggests a secondary binding site distinct from the
classical ANS binding site at the heme pocket. This also
points to the existence of a minor pathway for myoglobin unfolding, where a surface for dye binding is
exposed at a place other than the heme-binding pocket,
creating an additional binding site for ANS. The mass
spectral measurements avoid interference from the
heme, which is an unavoidable feature of optical spectroscopic measurements.
In fact, H–D exchange experiments on myoglobin
provided strong proof of the floppy nature of the
protein, which undergoes rapid exchange with the
solvent [24]. Time resolved ESI-MS of the acid unfolding of holomyoglobin demonstrates that myoglobin
undergoes acid expansion following titration of the
histidine residues bound to the heme [30].
Triosephosphate Isomerase
Triosephosphate isomerase (PfTIM) used in this study
was cloned from Plasmodium falciparum and expressed
in E. coli. PfTIM is a homodimeric ␤8/␣8 barrel protein,
whose folding pathway has been studied in various
denaturants [31, 32]. PfTIM has been demonstrated to
form a molten globule state at pH 3.0. Fluorescence
studies indicate that the pH 3.0 form of PfTIM shows
strong ANS binding. PfTIM upon pH unfolding exhibits a shift in charge state distribution from higher m/z
values to lower m/z values. ANS binding studies carried
out with pfTIM (Figure 1 right panel) shows that the
protein indeed binds a large amount of ANS at this pH
where it is known to exist as a molten globule. The
protein binds up to six ANS molecules. Below pH 2.5,
the protein binds no ANS. ANS binding is also not
detectable at pH 5.0. At pH 2.0, the protein does not
bind ANS and the charge state distribution shifts to
very low m/z values suggesting exposure of all the
residues that are likely to be protonated at this pH.
Indeed, solution studies indicate that PfTIM does not
bind ANS at this pH. PfTIM does not yield an ESI-MS
spectrum at pH 7.0, presumably because it forms a
strong dimer, whose charge states lie outside the quadrupole mass range (3000 Da) of the instrument. It is
interesting to note that even at pH 3.0, some charge
states are observed at the high end of the quadrupole
(2700 –3000 m/z), which do not show any satellite peaks
arising from bound ANS molecules. Close inspection of
the charge states at pH 3.0 clearly shows a bimodal
charge state distribution for the monomeric protein.
There are two maxima; one at 1849 m/z and the other at
1341 m/z. Chowdhury et al. [5] attributed such a bimodal distribution in cytochrome c to be due to distinct
protein structural states, which are populated in solution, which then produce distinct charge state distribution in the electrospray mass spectrum.
Carbonic Anhydrase
Carbonic anhydrase is a 29 kDa monomeric zinc containing protein. Figure 2 left panel shows the mass
spectra of carbonic anhydrase incubated with 100 ␮M
ANS and recorded at pH values of 2.5, 6.0, and 3.0. At
pH 2.5, the protein appears to be unfolded and yields a
mass of 29,020 Da, which corresponds to the mass of the
protein without the zinc. Indeed, the zinc is held by four
histidines (His3, His94, His96, His119), which are protonated at low pH leading to the release of zinc. At pH
6.0, the protein exhibits a charge state distribution that
is shifted towards the higher m/z values and yields a
mass of 29,083 Da, indicating that the protein is in its
zinc bound form and is compacted sufficiently. At pH
5.0, the protein undergoes mild acid expansion, although much of the protein still remains in the holo
form (not shown). The holo form did not bind any ANS
at pH 6.0 and 5.0. At pH 4.0, some charge states
corresponding to an ANS bound form were detected.
At pH 3.5, the protein binds up to four ANS molecules
per protein molecule; the single ANS bound form being
the most predominant.
It is interesting to note that the onset of ANS binding is
correlated to the release of zinc from the protein. Below
pH 3.0, very little ANS binding was detected. Similar
behavior is observed in studies of carbonic anhydrase
binding to inhibitors when monitored by ES-MS [33].
Cytochrome c
Cytochrome c is probably one of the best characterized
proteins both in solution and in the gas phase [14 –17,
34 –37] because of its small size and relative ease of
ionization. Different conformations are known to coexist in the absence of solvent [17].
Binding of ANS to bovine cytochrome c monitored
by ESI-MS, when incubated at different pH is shown in
the right panel of Figure 2. At pH 2.0, no ANS binding
is observed and the spectrum deconvolutes to the
molecular mass of free cytochrome c (12,231 Da). At pH
2.5, four sets of charge states are observed, one of which
is the free protein. The other charge states correspond to
one, two, and three ANS bound forms, respectively. At
pH 3.0, at least five different ANS bound species are
detectable, the species with two molecules of bound
ANS being the most abundant. At pH 4.0, only one ANS
bound species is observed and its abundance in solution
is drastically reduced when compared with that at pH
3.0. At pH 5.0, most of the protein exists in free form with
only a very small fraction still bound to ANS. At pH 6.0,
the ⫹8 charge state is the most intense peak. The folded
form of cytochrome c is known to give predominantly
charged states ⫹7 to ⫹9 in solution, whereas for unfolded
states the charge states range from ⫹12 to ⫹21 [5, 9].
These results suggest, the presence of a “molten
globule like” conformation with exposed hydrophobic
patches for cytochrome c at pH 3.0 as evidenced by its
enhanced affinity for ANS. Notably, molten globule like
states for cytochrome c have been implicated only in
mixed aqueous-organic solvents [46], or under membrane mimetic conditions [47]. The use of other spectroscopic probes suggests a cooperative unfolding transition with decrease in pH, in aqueous medium
containing a low concentration of methanol (3%). It is
conceivable that the low pH “unfolded states” of cytochrome c at pH 3.0 may still possess sufficient structure
to permit dye binding.
Lysozyme
The folding of lysozyme has been extensively studied. It
is a small structurally robust monomeric protein (14,305
Da). Much of its stability is due to the presence of the
Figure 2. ESI mass spectrum of carbonic anhydrase (left) and cytochrome c (right) at different pH
values in the presence of excess ANS. The masses obtained upon deconvolution are shown to the right
of each spectra. For carbonic anhydrase spectra are shown for pH values 2.5, 3.0, and 6.0, whereas for
cytochrome c, spectra at pH 2.0, 2.5, 3.0, and 4.0 are shown.
four disulfide crosslinks. Reduction of these disulfides
leads to dramatic destabilization of the protein [38].
Acid unfolding of the oxidized and reduced forms in
the presence of ANS monitored by ESI-MS spectra
clearly revealed major structural differences (Figure 3
right panel). Oxidized lysozyme (the native protein),
exhibited very little shift in charge states and did not
bind ANS at any pH. The observed charge states (⫹8 to
⫹13) is much less than the number of putative charge-
able sites on the protein (18 positively charged residues ⫹ the free N-terminus). This suggests that even at
low pH the protein remains substantially compact and
structured. Upon reduction with 200 mM DTT, the
spectra becomes distinctly rich in charge states with
states bearing up to 21 positive charges appearing at pH
2.0. The reduced form did not bind ANS between pH
2.0 and 5.0. At pH 7.0, it shows a bimodal distribution
of charge states, which deconvolute to 14,314.7 Da
Figure 3. ANS binding to oxidized and reduced forms of ␣ lactalbumin (left) and lysozyme (right):
(A1) Oxidized ␣-lactalbumin at pH 7.0. (B1) Oxidized ␣-lactalbumin at pH 3.0. (C1) Reduced
␣-lactalbumin at pH 3.0. (D1) Reduced ␣-lactalbumin at pH 7.0. (A2) Oxidized lysozyme at pH 7.0.
(B2) Oxidized lysozyme at pH 3.0. (C2) Reduced lysozyme at pH 3.0. (D2) Reduced lysozyme at pH
7.0. All spectra were recorded after incubation with ANS.
(weight of the reduced form). The spectrum has two
maxima, one at 1591.4 m/z (⫹9) and the other at 1021.4
(⫹14), suggesting that the protein after reduction and
incubation at pH 7.0, is partially folded and exists in
equilibrium between multiple conformations. Under
these conditions, lysozyme bound up to three ANS
molecules with the molecular weights 14,614.74 Da
(lysozyme ⫹ 1ANS), 14,914 Da (lysozyme ⫹ 2ANS)
and 15,212 Da (lysozyme ⫹ 3ANS) observed upon deconvolution. Many fluorescence studies in solution
have shown that reduced lysozyme indeed binds ANS
at near neutral pH [39].
␣-Lactalbumin
Although structurally similar to lysozyme including the
presence of four disulfides, ␣-lactalbumin (14,177 Da)
has a different pH unfolding profile. It is known to form
a molten globule state called the A state at low pH [40].
Other methods of inducing molten globules in ␣-lactalbumin include reduction and calcium removal [40].
Unlike lysozyme, four ANS bound complexes were
detected with oxidized ␣-lactalbumin at pH 3.0 (Figure
3), the most intense being the two ANS bound form. A
modest shift in the charge state distribution was observed, suggesting acid expansion of the molecule.
Reduction with DTT leads to a marked increase in the
number of detectable charge states. Between pH 2.0 and
4.0, no ANS binding is observed with reduced ␣-lactalbumin (Figure 3). At pH 5.0, one bound ANS molecule
was detected and at pH 7.0, two ANS molecules were
bound (Figure 3 D1). Previously reported spectroscopic
studies on reduced forms of ␣-lactalbumin containing
one to three disulfides (3SS, 2SS, and 1SS), indicate that
all of them appear to be molten globule like structures,
with exposed hydrophobic surfaces, which can bind
ANS [41].
BPTI
BPTI is a small monomeric 6 kDa protein, which acts as
an inhibitor of bovine trypsin. The protein is stabilized
by three disulfide bridges, and the reduction of these
disulfide bridges leads to loss of structure. Extensive
folding studies have been carried out on BPTI [42].
Mutants of BPTI lacking disulfide bonds form molten
globules and show enhancement of ANS fluorescence
indicative of dye binding [43, 44]. ESI-MS studies established that BPTI undergoes acid denaturation with a
monotonous shift in charge state distribution, without
any evidence for any unfolding intermediates. Like
lysozyme, oxidized BPTI does not bind ANS across the
pH range of 2.0 to 7.0 as monitored by ESI-MS. The
reduced form exhibits a marked shift in the charge state
distribution towards lower m/z values. Reduced BPTI
bound a single ANS molecule at near neutral pH and
none at acidic pH (data not shown) suggesting the
presence of a floppy structure at pH 7.0 with not many
hydrophobic sites competent for ANS binding.
Effect of Charge States on ANS Binding
The results presented above suggest that there is a
strong correlation between ANS binding observed in
the gas phase and the structural state of the protein in
solution. The association of ANS with expanded, molten globule states of proteins in solution, is maintained
under mass spectrometric conditions. The mild ionization conditions facilitate the maintenance of noncovalent interactions that promote dye binding to proteins.
In the case of protein molten globule states, it is very
difficult to establish stoichiometry of ANS binding in
solution using conventional fluorescence methods due
to the presence of multiple equilibria and binding site
heterogeneity. In proteins like myoglobin and cytochrome c, the presence of the heme hampers optical
spectroscopic measurements. ESI-MS provides a direct
measurement of the number of ANS molecules bound
per protein molecule. For example, smaller proteins like
myoglobin (6 ANS) and cytochrome c (5 ANS) bind a
larger number of ANS molecules than a larger protein
like carbonic anhydrase (4 ANS). The binding of ANS
thus appears to be a property of protein structure rather
than the size.
The ready detection of multiple dye-protein complexes in the gas phase suggests that ESI-MS may be
used to determine the distribution of various species in
solution, assuming that all species ionize with equal
efficacy under electrospray conditions. This assumption
however may not be generally valid in all cases [1].
Close inspection of the intensities of the satellite peaks
around each charge state, suggests that individual
charge states have different affinities for ANS. It was
observed for cytochrome c and myoglobin (data not
shown) that the lower charge states have one or two
ANS bound species as the major components, whereas
in the higher charge states, the free protein dominates.
In the case of cytochrome c, the ⫹7 charge state has
the single ANS bound species as the major component, whereas the ⫹13 charge state has the free
protein as the dominant species. Similarly, in the case
of myoglobin, the ⫹9 charge state has the single ANS
bound species as the most intense component,
whereas for the ⫹17 charge state, the free protein
peak is most intense.
In the case of PfTIM, the charge states are complicated by the fact that at pH 3.0, there are probably
multiple conformations in equilibrium (Figure 4A). The
expanded regions corresponding to four charge states
are shown (Figure 4B–E). The charge state ⫹13 has very
few low intensity satellite peaks due to bound ANS. The
⫹16 charge state shows several ANS molecules bound
to it with the one ANS bound species being the most
intense. The ⫹24 charge state shows reduced ANS
binding but still has the single ANS bound component
as the dominant species. The charge state ⫹27 has the
free protein as the most dominant species. The possibility that ANS binding is dependent on the overall
resident positive charge of the protein may be discounted since both the lower charge state ⫹13 (Figure
4B) and the higher charge state ⫹27 (Figure 4C) show
diminished ANS binding in the case of PfTIM, suggesting, that the observed binding may in fact be a property
of the structural state of the protein. Indeed, hydrogen–
deuterium exchange studies carried on gas phase ions
of cytochrome c [35], lysozyme, and BPTI [36] support
Figure 4. (A) ESI-MS spectrum of pfTIM (15 ␮M) incubated with excess ANS (100 ␮M) recorded at
pH 3.0. (B–E) Expanded regions of this spectrum showing charge states ⫹13, ⫹16, ⫹24, and ⫹27,
respectively.
the view that the gas phase conformations are different
for specific charge states. Furthermore, ion mobility
studies by Clemmer et al., on lysozyme ions indicate
that lower charge states have different drift times as
compared to higher charge states [38]. The higher
charge states have been proposed to have an elongated
structure as compared to lower charge states; a feature
related to the greater internal energies of gas phase
protein ions [38]. Ions with greater internal energy have
been shown to form poor noncovalent complexes with
ligands. Therefore, the effect of increasing the internal
energy of the protein molecule on the distribution of
ANS can be investigated by increasing the nozzle
skimmer potential.
Figure 5. ESI mass spectra of ANS adducts from myoglobin (15 ␮M) incubated with ANS (100 ␮M)
and recorded at different nozzle skimmer potential (⌬V). Spectra obtained at 30, 60, 75, and 100 V are
shown.
Effect of Nozzle Skimmer Potential
The skimmers are located in the collision zone of the
mass spectrometer at the ion sampling interface, which
accelerates the ions and makes them collide with the
collision gas (nitrogen in this case); this causes an
increase in the internal energy of the molecule, which in
turn affects conformation of protein ions. For example,
in the case of myoglobin, Douglas and co-workers have
reported that myoglobin can be unfolded partially in
the gas phase by varying nozzle skimmer potentials
[45]. Myoglobin and cytochrome b5 ions expand in size
and then lose the heme, giving rise to the apo forms
[45]. ANS bound noncovalent complexes were subjected to increasing nozzle skimmer potential to study
their relative stabilities. Figure 5 shows the spectra
obtained for dissociation of ANS from apomyoglobin as
a function of skimmer potential. As expected, it was
found that with increasing nozzle skimmer potential
the ANS molecules were systematically stripped from
the protein. ANS dissociates first from species bearing
multiple molecules of ANS. At relatively high skimmer
potentials, only single ANS bound species and unbound protein are detectable. Douglas and co-workers
[45] have used these potential values to measure gas
phase binding energies. For comparison, dissociation of
heme from myoglobin was also carried out as a function
of nozzle skimmer potential. No attempt was made in
the present study to estimate the gas phase binding
energy of protein–ANS complexes, because under the
current experimental conditions instrumental parameters like drift time, collision cross section area, and gas
number density at the sampling orifice could not be
determined. Nevertheless a qualitative comparison of
heme and ANS dissociation is discussed. Figure 6A
shows a plot for dissociation of ANS from apomyoglobin and heme from myoglobin as a function of skimmer
potential. Similar plots were obtained for cytochrome c,
carbonic anhydrase, and lysozyme to study the relative
strengths of ANS bound species. The dissociation plots
for ANS from these molecules as a function of nozzle
skimmer potential is shown in Figure 6B. Although it is
difficult to directly compare dissociation profiles obtained for ANS binding to proteins of widely differing
masses, some qualitative features may be inferred from
the data. ANS is easily dissociated from complexes with
the reduced forms of lysozyme at very low nozzle
skimmer potential suggesting, a relatively weak association between the dye and the macromolecule. Similar
results have been obtained with reduced forms of
␣-lactalbumin and BPTI suggesting that these proteins
do not have appreciable structure in the absence of
disulfide bonds resulting in only weak binding of the
dye. It is anticipated that the dye complexes of larger
proteins will require higher dissociation voltages precluding a direct comparison with significantly smaller
macromolecules. Nevertheless, the data in Figure 6B
shows significant differences in dissociation voltages of
the proteins triosephosphate isomerase (M r ⫽ 27,831
Figure 6. (a) Dissociation curves of ANS and heme from myoglobin as a function of skimmer potential. (b) Dissociation curves
showing fraction of ANS bound to different proteins as a function
of skimmer potential. The fraction of dye bound to the protein was
determined by averaging over the total ion chromatogram and
using the intensities of the species obtained at different masses in
the deconvoluted spectrum. The fraction bound includes all the
multiply bound forms.
Da) and carbonic anhydrase (M r ⫽ 29,021 Da) suggesting that ANS binding is significantly stronger in the
former. It is pertinent to note that Hunter et al. [45],
observed that although cytochrome b5 (M r ⫽ 15,198
Da) and myoglobin (M rapo ⫽ 16,950 Da) have similar
binding affinities for heme binding in solution, the gas
phase higher skimmer potentials are necessary for the
dissociation of heme from myoglobin than cytochrome
b5. Inspection of the charge states of the protein–ANS
complexes as a function of nozzle skimmer potential
suggests that the higher charge states lose the bound
ANS before the lower charge states. The higher skimmer potentials probably affect higher charge state structures more than lower charge states causing release of
ANS. The pattern dissociation of ANS from carbonic
anhydrase was found to be similar to the dissociation of
ligand complexes of carbonic anhydrase as reported
recently by Gao et al. [46].
Conclusions
The present study demonstrates the utility of ESI-MS as
a rapid and a sensitive tool to detect ANS complexes in
the gas phase. For many proteins maximal ANS binding
is observed in the pH range 2.5– 4.0 where molten
globules have been demonstrated to exist. It is likely
that these acid expanded partially unfolded states provide greater surface area for dye binding. The protein–
dye complexes do not dissociate under electrospray
conditions. The stoichiometries of dye bound complexes can thus be easily established. The robustness of
these adducts in the gas phase suggests that electrostatic interactions between the sulfonate moiety of the
dye and protein may be important determinants of the
stability of these complexes. Although a great body of
literature suggests that ANS binds at hydrophobic
regions, recent work has emphasized on the electrostatic nature of ANS binding [20]. For a given charge
state distribution, binding is observed to be dependent
on the net charge on a given ion. No correlation is found
between the number of ANS bound complexes and
protein size. Carbonic anhydrase (29,021 Da) binds only
three, whereas cytochrome c (12,230 Da) binds as many
as five ANS molecules.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
Acknowledgments
29.
The mass spectrometry facility is funded by the Department of
Biotechnology (Govt. of India) as program support to the area of
“Drug and Molecular Design.” SKS acknowledges receipt of a
senior research fellowship from the Council of Scientific and
Industrial Research (CSIR), Govt. of India.
30.
References
33.
1. Loo, J. A. Mass Spectrom. Rev. 1997, 16, 1–23.
2. Veenstra, T. D. Biophys. Chem. 1999, 79, 63–79.
3. Smith, D. L.; Deng, Y.; Zhang, Z. J. Mass Spectrom. 1997, 32,
135–146.
4. Cherokee, S.; Hoaglund-Hyzer •.; Counterman, A. E.; Clemmer, D. E. Chem. Rev. 1999, 3037–3080.
5. Chowdhury, S. K.; Katta, V.; Chait, B. T. J. Am. Chem. Soc. 1990,
112, 9012–9013.
6. Chung, E. W.; Nettleton, E. J.; Morgan, C. J.; Gross, M.;
Miranker, A.; Radford, S. E.; Dobson, C. M.; Robinson, C. V.
Protein Sci. 1997, 6, 1316 –1324.
7. Katta, V.; Chait, B. T. J. Am. Chem. Soc. 1991, 113, 8534 – 8535.
8. Konermann, L.; Douglas, D. J. Rapid Commun. Mass Spectrom.
1998, 12, 435– 442.
9. Konermann, L.; Collings, B. A.; Douglas, D. J. Biochemistry
1997, 36, 5554 –5559.
10. Ray, S. S.; Balaram, P. J. Phys. Chem. B 1999, 103, 7068 –7072.
11. Weber, G.; Teals, F. W. J. Trans. Faraday. Soc. 1957, 53, 656.
12. Slavic, J. Biochim. Biophys. Acta. 1982, 694, 1–25.
13. Wolynes, P. G. Proc. Natl. Acad. Sci. USA 1995, 92, 2426 –2427.
14. Wagner, D. S.; Andregg, R. J. Anal. Chem. 1994, 66, 706 –711.
15. McLafferty, F. W.; Guan, Z.; Haupts, U.; Wood T. D.;
Kellecher, N. L. J. Am. Chem. Soc. 1998, 120, 4732– 4740.
16. Ogorzalek, R. R.; Loo, J. A.; Dales, N.; Andrews, P. C. Protein
Sci. 1994, 3, 1975–1983.
17. Sukau, D.; Shi, Y.; Beu, S. C.; Senko, M. W.; Quinn, J. P.;
31.
32.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
Wampler F. M., III; McLafferty, F. W. Proc. Natl. Acad. Sci. USA
1993, 90, 790 –793.
Eckart, K.; Verhey, J.; Spiess, J. Proceedings of 44th ASMS
Conference on Mass Spectrometry and Allied Topics., 1996; p 1066.
McLuckey, S. A.; Stephenson, J. L., Jr. Mass Spectrom. Rev. 1998,
17, 369 – 407.
Ali, V.; Prakash, K.; Kulkarni, S.; Ahmad, A.; Madhusudhan,
K. P.; Bhakuni, V. Biochim. Biophys. Acta 1999, 1432, 194 –202.
Matulis, D.; Baumann, C. G.; Bloomfield, V. A.; Lorien, R. E.
Biopolymers 1999, 49, 451– 458.
Ranie, J.; Kumar, V. P; Balaram, H. Mol. Biochem. Parasitol.
1993, 61, 159 –169.
Feng, R.; Konishi, Y. J. Am. Soc. Mass Spectrom. 1993, 4,
638 – 645.
Wang, F.; Tang, X.-j. Biochemistry 1996, 35, 4069 – 4078.
Konishi, Y.; Feng, R. Biochemistry 1994, 33, 9706 –9711.
Stephenson, J. L., Jr.; Van Berkel, G. J.; McLuckey, S. A. J. Am.
Soc. Mass Spectrom. 1997, 8, 637– 644.
Fontana, A.; Zambonin, M.; Polverino de Laureto, P; De
Filippis, V; Clementi, A.; Scaramella, E. J. Mol. Biol. 1997, 266,
223–230.
Tsui, V.; Garcia, C.; Cavagnero, S.; Suizdak, G.; Dyson, H. J.;
Wright, P. E. Protein Sci. 1999, 8, 45– 49.
Hughson, F. M.; Wright, P. E.; Baldwin, R. L. Science 1990, 249,
1544 –1548.
Konermann, L.; Rossel, F. I.; Mauk. A. G.; Douglas. D. J.
Biochemistry 1997, 36, 6448 – 6454.
Gokhale, R. S.; Ray, S. S.; Balaram, H.; Balaram, P. Biochemistry
1999, 38, 423– 431.
Gopal, B.; Ray, S. S.; Gokhale, R. S.; Balaram, H.; Murthy,
M. R. N.; Balaram, P. Biochemistry 1999, 38, 478 – 486.
Wu, Q.; Gao, D.; McCarthy, J.; Sigal, G. D.; Bruce, J. E.;
Whitesides, G. M.; Smith, R. D. J. Am. Chem. Soc. 1997, 119,
1157–1158.
Clemmer, D. E.; Hudgins, R. R.; Jarrold, M. F. J. Am. Chem. Soc.
1995, 118, 4488 – 4489.
Valentine, S. J.; Clemmer, D. E. J. Am. Chem Soc. 1996, 119,
3558 –3566.
Shelimov, K. B.; Clemmer, D. E.; Hudgins, R. R.; Jarrold, M. F.
J. Am. Chem. Soc. 1997, 119, 2240 –2248.
Schnier, P. D.; Gross, D. S.; Williams, E. R. J. Am. Chem. Soc.
1995, 117, 6747– 6757.
Valentine, S. J.; Anderson, J. G.; Ellington, A. D.; Clemmer,
D. E. J. Phys. Chem. B 1997, 101, 3891–3900.
Itzhaki, L. S.; Evan P. A.; Dobson, C. M.; Radford, S. E.
Biochemistry 1994, 33, 5212–5220.
Kuwajima, K. FASEB J. 1996, 10, 102–109.
Mac, L. C.; Anderson, S. Biochemistry 1997, 36, 3728 –3732.
Creighton, T. E.; Darby, N. J.; Kemmink, J. FASEB J. 1996, 10,
110 –118.
Ferrer, M.; Barany, G.; Woodward, C. Natl. Struct. Biol. 1995, 2,
211–217.
Dadlez, M. Biochemistry 1997, 36, 2788 –2797.
Hunter, C. L.; Mauk, A. G.; Douglas, D. J. Biochemistry 1997, 36,
1018 –1025.
Konerman, L.; Douglas, D. J. Biochemistry 1997, 36, 12296 –
12302.
Bychkova, V. E.; Dujsekina, A. E.; Klenin, S. I.; Tiktopulo, E. I.;
Uversky, V. N.; Ptitsyn, O. B. Biochemistry 1996, 35, 6058 – 6063.